Yesterday, in Madrid, the Council of Ministers approved the bill for "the proper use and governance of artificial intelligence." This regulation aims to ensure an ethical, inclusive, and beneficial use of AI, penalizing certain practices with fines up to 35 million euros or 7% of a company's annual turnover. Among these practices is the absence of labeling on any image, audio, or video generated or manipulated by AI.
The bill aligns with the directives of the AI Act which imposes strict transparency obligations on AI systems considered high-risk. Oscar Lopez, Minister of Digital Transformation and Public Administration, emphasized during the press conference following the Council of Ministers that "Spain is one of the first countries in the European Union to apply these rules, considered more comprehensive than those of the United States, which largely rely on voluntary compliance and a series of state regulations."
In addition to the absence of labeling on AI-generated content, considered a serious offense, the legislation also targets other at-risk practices. Among them, the use of subliminal techniques aimed at manipulating vulnerable groups. Oscar Lopez notably cited the example of chatbots identifying gambling-addicted users and exploiting this vulnerability to encourage further gambling. He also mentioned interactive toys encouraging children to take on dangerous challenges.
Furthermore, the law prohibits biometric classification based on ethnicity, political opinions, religion, or sexual orientation. Social scoring, which evaluates individuals based on behaviors or personal characteristics, is banned when used to restrict access to subsidies, deny a loan, or predict the likelihood of a person committing a crime based on personal data, such as family background, place of residence, or education level.
The enforcement of these new rules will fall under the Spanish Agency for AI Oversight (AESIA), the first European public authority dedicated exclusively to the regulation and supervision of artificial intelligence. However, some areas will be subject to sectoral oversight: the Spanish Data Protection Agency (AEPD) will handle issues related to biometrics, the Central Electoral Board will oversee AI systems likely to influence the democratic process, while the General Council of the Judiciary will be responsible for those impacting the administration of justice.